Dr Piotr Kijanka at the 2024 IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium
Dr Piotr Kijanka, professor at the Faculty of Mechanical Engineering and Robotics, ranked first and third in two categories at a-Mem Challenge 2024 which evaluates the quality of the developed algorithms used to estimate elastic modulus in ultrasonic elastography. The competition has seen participants of the international “2024 IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium,” held on 22–26 September 2024 in Taipei (Taiwan).
“Participants in the competition received raw measurement data collected from various soft materials in two categories: propagating shear waves from impulse sources with a focused ARF beam (acoustic radiation force) and propagating shear waves from harmonic (continuous) sources. The task was to create two-dimensional elasticity maps for both homogeneous and heterogeneous materials,” reported Professor Piotr Kijanka.
The AGH University scholar took the 1st place in harmonic elastography, whereas in impulse elastography he was ranked third.
“The biggest difficulty in both categories was adapting the algorithms to work as optimally as possible for all the data provided. Due to the variety of materials tested, both homogeneous and heterogeneous, it was crucial to develop a solution that would accurately estimate the elastic modulus in all cases while effectively detecting and analysing the propagating transverse waves.”
Dr Piotr Kijanka during the IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium. Photograph: private archive
In ultrasound elastography, ultrasonic waves which penetrate into the body are used as a probe. Organs subjected to impulses of a focused acoustic beam begin to gently vibrate and emit elastic transverse waves registered by the device. The technique takes advantage of the fact that pathological processes cause gradual changes in the mechanical properties of organs. Diseased organs subjected to an impulse of a focused acoustic beam vibrate differently from healthy organs, which translates into a change in the speed and profile of propagation of transverse waves. If they deviate from the reference values, this is a signal that something wrong is happening in the patient's body.
The development of more accurate algorithms used in ultrasonic elastography can help, among other things, to overcome the problem of reproducibility of tests obtained with this technique resulting from different equipment configurations or different measurement standards used by equipment manufacturers.
“Another benefit is increased precision in detecting and differentiating pathological changes. More accurate algorithms identify subtle differences in tissues better, which is particularly relevant in the diagnosis of neoplasms, fibrosis, and other diseases of organs such as liver and pancreas. Doctors can more quickly and accurately assess the stage of the disease and differentiate between benign and malignant lesions, which reduces the time from the diagnosis to the implementation of the right treatment,” explained Professor Kijanka. “The reliability of the measurements also improves the ability to monitor disease progression and the effectiveness of therapy over the long term. Owing to accurate and reproducible results, doctors can assess whether a patient is responding to treatment with greater certainty, which allows to dynamically adjust the therapy. It is particularly valuable in case of chronic diseases such as liver fibrosis, as the state of the tissue may change in response to treatment."
More accurate algorithms may also obviate the need for invasive examinations in the future, which in some cases may be replaced by ultrasound elastography.
The idea behind the Algorithms for Mapping Elastic Modulus (a-MEM) Challenge 2024 is to challenge the community to devise more accurate and precise algorithms for elastic modulus estimation, raise greater awareness of the challenges associated with elastic modulus estimation, and gain further knowledge on why certain approaches fail. The organisers were also guided by the strive to identify the most promising approaches to elastic modulus estimation and provide a benchmark and public data for the future development of SWS and elastic modulus estimation algorithms.